The ABINIT Users Mailing List ( CLOSED )

["D. R. Hamann" <drh@bell-labs.com>]

Dear Martin,

This was caused by a bug that is rather simple to fix.  Delete lines 330
and 347 in Src_6response/nselt3.f, and recompile.  I'm attaching an
excerpt from this routine with these lines indicated.

Good luck!

Don Hamann

Martin Tijssens wrote:

Dear abinit-users,

I have a problem calculating elastic constants of NiTi. I did a
calculation on a pentium IV and SGI R10000, both with abinis and abinip,
compiled by me with IFC and PG compilers, but both calculations end with
the following message:


 ETOT 78   1.3863424114599    -1.336E-12 5.275E-27 1.227E-21
 At SCF step   78       vres2   =  1.23E-21 < tolvrs=  1.00E-20 =>converged.
lib-4201 : UNRECOVERABLE library error
Encountered during a sequential unformatted READ from unit 17
Fortran unit 17 is connected to a sequential unformatted  file:
  "NiTitmp_KG"
IOT Trap


The calculation stops as the first perturbation in dataset 4 (see below)
has almost converged. Can anybody shed some light on this?

I use the following input file:

# NiTi strain perturbation
 ndtset   4
 mkmem    0
 mkqmem   0
 mk1mem   0
 mffmem   0
# Set 1 : initial self-consistency
  iscf1     5
  kptopt1   1
  prtden1   1
  tolvrs1   1.0d-20
#Dataset 2: non-SCF for all k points
  getden2   1
  kptopt2   3
  getwfk2   1
    iscf2  -2
  tolwfr2   1.0d-20
#Dataset 3: non-SCF calculation
   getden3  1
   getwfk3  2
   kptopt3  3
     nqpt3  1
     iscf3 -2
      qpt3  0.0d0 0.0d0 0.0d0
   tolwfr3  1.0d-20
#Dataset 4: RF calculation
   diemix4  0.5
   getden4  1
   getwfk4  2
   kptopt4  3
     nqpt4  1
   getwfq4  3
      qpt4  0.0d0 0.0d0 0.0d0
  rfatpol4  1 2
    rfdir4  1 1 1
   rfstrs4  3
   tolvrs4  1.0d-20
     iscf4  3
# common input data
   ecutsm   0.0
     acell   3*5.718
     rprim   1 0 0   0 1 0   0 0 1
    occopt   4
    tsmear   0.02
    ntypat   2
     znucl   22.0 28.0
     natom   2
     typat   1 2
      xred   0.0 0.0 0.0 0.5 0.5 0.5
     nband   18
      ecut   36
     nshiftk 1
     shiftk  0.5 0.5 0.5
     ngkpt   12 12 12
     nstep   200

  

-- 
D. R. Hamann				Phone: 908-582-4454
Director, Theoretical Materials    	Fax:   908-582-4702
  Physics Research (retired)		email: drh@physics.bell-labs.com
Bell Laboratories 
Lucent Technologies
700 Mountain Ave, Room 1D-371
Murray Hill, NJ 07974-0636 USA

*********** From Src_6response/nselt3.f  in abinit_4.3.3  ********


   if (mkmem==0) then
!   Read (k+G) basis sphere data (same for each spin)
    call rdnpw(ikpt,isppol,nband_k,npw_k,nspinor,0,unkg)

!   Read sphere data centered at k in unkg, then k+g data
    read (unkg)  !**************************************** delete line 330
    read (unkg) ((kg_k(ii,muig),ii=1,3),muig=1,npw_k)

   else

    kg_k(:,1:npw_k)=kg(:,1+ikg:npw_k+ikg)

!  End if for choice governed by mkmem
   end if

   wtk_k=wtk_rbz(ikpt)

   if (mk1mem==0) then
!   Read (k+q+G) basis sphere data (same for each spin)
    call rdnpw(ikpt,isppol,nband_k,npw1_k,nspinor,0,unkg1)

!   Read sphere data centered at k in unkg, then k+g data
    read (unkg1)  !**************************************** delete line 347
    read (unkg1) ((kg1_k(ii,muig),ii=1,3),muig=1,npw1_k)

   else